langium-tools 0.3.2 → 0.4.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
package/README.md
CHANGED
|
@@ -28,12 +28,13 @@
|
|
|
28
28
|
A collection of tools for implementing and testing DSLs with <a href="https://langium.org/"><strong>Langium</strong>!</a>
|
|
29
29
|
<br />
|
|
30
30
|
<br />
|
|
31
|
-
<a href="
|
|
31
|
+
<a href="https://langium-tools.openbrod.org"><strong>Website »</strong></a>
|
|
32
32
|
<br />
|
|
33
|
-
<a href="https://
|
|
33
|
+
<a href="https://langium-tools.openbrod.org/api"><strong>API reference »</strong></a>
|
|
34
|
+
<br />
|
|
35
|
+
<a href="#documentation"><strong>Documentation in this README »</strong></a>
|
|
34
36
|
<br />
|
|
35
37
|
<br />
|
|
36
|
-
|
|
37
38
|
<a href="https://github.com/borisbrodski/langium-tools/issues/new?labels=bug&template=bug-report---.md">Report Bug</a>
|
|
38
39
|
·
|
|
39
40
|
<a href="https://github.com/borisbrodski/langium-tools/issues/new?labels=enhancement&template=feature-request---.md">Request Feature</a>
|
|
@@ -109,7 +110,7 @@ This section should list any major frameworks/libraries used to bootstrap your p
|
|
|
109
110
|
|
|
110
111
|
- Add package to your project
|
|
111
112
|
- Start using one or more features it provides
|
|
112
|
-
-
|
|
113
|
+
- Read the [state machine showcase](https://github.com/borisbrodski/langium-tools-showcases/tree/main/statemachine) for a full language that uses the library
|
|
113
114
|
|
|
114
115
|
### Installation
|
|
115
116
|
|
|
@@ -123,13 +124,35 @@ Add the package to your `package.json`
|
|
|
123
124
|
|
|
124
125
|
### Example project
|
|
125
126
|
|
|
126
|
-
|
|
127
|
+
The [state machine showcase](https://github.com/borisbrodski/langium-tools-showcases/tree/main/statemachine) is the official Langium statemachine. The grammar and the C++ program are the original example. langium-tools replaces the hand-written parts around them. `git diff upstream -- statemachine` in that repository is the extension.
|
|
128
|
+
|
|
129
|
+
Clone it next to this repository. The example depends on `file:../../langium-tools`, and this package exports `dist/`, so build here first:
|
|
130
|
+
|
|
131
|
+
```sh
|
|
132
|
+
git clone https://github.com/borisbrodski/langium-tools.git
|
|
133
|
+
git clone https://github.com/borisbrodski/langium-tools-showcases.git
|
|
134
|
+
cd langium-tools && npm install && npm run build
|
|
135
|
+
cd ../langium-tools-showcases/statemachine && npm install && npm test
|
|
136
|
+
```
|
|
127
137
|
|
|
128
|
-
|
|
138
|
+
What that language uses, and why it is less code:
|
|
139
|
+
|
|
140
|
+
- `getDocumentIssueSummary` is the CLI's only error report. Lexer, parser, and validation errors share one count. Warnings are printed and generation continues. The old loop kept diagnostics of severity 1 and could miss a syntax error.
|
|
141
|
+
- `toHaveNoErrors` ignores warnings. `toHaveNoIssues` does not. A lowercase state name is the test that shows the difference, so a style warning does not have to fail a parse test.
|
|
142
|
+
- `parseWithMarks` and `toHaveDocumentIssues` use `<<{|` … `|}>>` as the expected range. The matcher fails when an issue is missing and when an extra issue appears. Hard-coded ranges such as `[1:19..1:26]` do not.
|
|
143
|
+
- `generate` only calls `createFile`. The CLI writes the directory. Identical bytes keep their modification time. The same path with different text throws before a file is overwritten. The C++ templates themselves are unchanged.
|
|
144
|
+
- `npm test` verifies `generated/` and `generated-metadata.json`. `npm run test:generate` rewrites them. The mode is `GENERATOR_TEST`. The C++ no longer has to be pasted into the test source.
|
|
145
|
+
- `JavaImportManager` turns the events into a Java enum. `useClass('java.util.function.Consumer')` prints `Consumer` and adds the import. A second type with that simple name stays fully qualified. Classes in the current package are not imported.
|
|
146
|
+
|
|
147
|
+
`generateJavaFile` is the one-call wrapper around the import manager. It builds the file with generator nodes from this package's Langium install. A project that also depends on Langium has a second copy, and those nodes print as `[object Object]`. The showcase calls the import manager and `createFile` instead. The note is in `statemachine/src/cli/generator.ts`.
|
|
148
|
+
|
|
149
|
+
`arraySortBy`, `adjustedUnix`, and named generator targets are real features. The state machine does not use them: event order is the language order, the snapshots are `\n`, and the `.cpp` and `.java` share one output directory. Named targets are the next step when those two files must land in different directories.
|
|
129
150
|
|
|
130
151
|
## Documentation
|
|
131
152
|
|
|
132
|
-
|
|
153
|
+
Guides for every part of the library, with examples, are on the website: [langium-tools.openbrod.org](https://langium-tools.openbrod.org). The [showcases](https://langium-tools.openbrod.org/showcases/) page is the index of working languages, starting with the state machine.
|
|
154
|
+
|
|
155
|
+
The generated API reference is [typedoc](https://langium-tools.openbrod.org/api).
|
|
133
156
|
|
|
134
157
|
### Document Issues
|
|
135
158
|
|
|
@@ -234,13 +257,13 @@ The module defines the following enums for issue sources and severities:
|
|
|
234
257
|
|
|
235
258
|
For detailed API documentation, see the Typedoc documentation.
|
|
236
259
|
|
|
237
|
-
- [getDocumentIssueSummary](https://
|
|
238
|
-
- [getDocumentIssues](https://
|
|
239
|
-
- [documentIssueToString](https://
|
|
240
|
-
- [DocumentIssue](https://
|
|
241
|
-
- [DocumentIssueSummary](https://
|
|
242
|
-
- [DocumentIssueSource](https://
|
|
243
|
-
- [DocumentIssueSeverity](https://
|
|
260
|
+
- [getDocumentIssueSummary](https://langium-tools.openbrod.org/api/functions/base.getDocumentIssueSummary.html)
|
|
261
|
+
- [getDocumentIssues](https://langium-tools.openbrod.org/api/functions/base.getDocumentIssues.html)
|
|
262
|
+
- [documentIssueToString](https://langium-tools.openbrod.org/api/functions/base.documentIssueToString.html)
|
|
263
|
+
- [DocumentIssue](https://langium-tools.openbrod.org/api/types/base.DocumentIssue.html)
|
|
264
|
+
- [DocumentIssueSummary](https://langium-tools.openbrod.org/api/types/base.DocumentIssueSummary.html)
|
|
265
|
+
- [DocumentIssueSource](https://langium-tools.openbrod.org/api/enums/base.DocumentIssueSource.html)
|
|
266
|
+
- [DocumentIssueSeverity](https://langium-tools.openbrod.org/api/enums/base.DocumentIssueSeverity.html)
|
|
244
267
|
|
|
245
268
|
#### Notes
|
|
246
269
|
|
|
@@ -379,10 +402,10 @@ await manager.writeToDisk(outputDir: string, target?: string, clean?: boolean);
|
|
|
379
402
|
|
|
380
403
|
For detailed API documentation, see the Typedoc documentation.
|
|
381
404
|
|
|
382
|
-
- [GeneratedContentManager](https://
|
|
383
|
-
- [GeneratorManager](https://
|
|
384
|
-
- [GeneratorTarget](https://
|
|
385
|
-
- [CreateFileOptions](https://
|
|
405
|
+
- [GeneratedContentManager](https://langium-tools.openbrod.org/api/classes/generator.GeneratedContentManager.html)
|
|
406
|
+
- [GeneratorManager](https://langium-tools.openbrod.org/api/interfaces/generator.GeneratorManager.html)
|
|
407
|
+
- [GeneratorTarget](https://langium-tools.openbrod.org/api/interfaces/generator.GeneratorTarget.html)
|
|
408
|
+
- [CreateFileOptions](https://langium-tools.openbrod.org/api/interfaces/generator.CreateFileOptions.html)
|
|
386
409
|
|
|
387
410
|
#### Notes
|
|
388
411
|
|
|
@@ -556,16 +579,16 @@ The generator testing utilities work by:
|
|
|
556
579
|
|
|
557
580
|
#### API Reference
|
|
558
581
|
|
|
559
|
-
For detailed API documentation, see the [Typedoc documentation](https://
|
|
582
|
+
For detailed API documentation, see the [Typedoc documentation](https://langium-tools.openbrod.org/api).
|
|
560
583
|
|
|
561
|
-
- [langiumGeneratorSuite](https://
|
|
562
|
-
- [GeneratorTestOptions](https://
|
|
563
|
-
- [GeneratedContentManager](https://
|
|
584
|
+
- [langiumGeneratorSuite](https://langium-tools.openbrod.org/api/functions/testing.langiumGeneratorSuite.html)
|
|
585
|
+
- [GeneratorTestOptions](https://langium-tools.openbrod.org/api/interfaces/testing.GeneratorTestOptions.html)
|
|
586
|
+
- [GeneratedContentManager](https://langium-tools.openbrod.org/api/classes/generator.GeneratedContentManager.html)
|
|
564
587
|
|
|
565
588
|
#### Notes
|
|
566
589
|
|
|
567
590
|
- **Environment Variable**: The `GENERATOR_TEST` environment variable controls the test mode (`verify` or `generate`).
|
|
568
|
-
- **Testing Framework**: The examples use `vitest`, but the utilities can be used with any testing framework, if you are not using automatic workspace discovery, but rather define one test for one workspace using [langiumGeneratorTest](https://
|
|
591
|
+
- **Testing Framework**: The examples use `vitest`, but the utilities can be used with any testing framework, if you are not using automatic workspace discovery, but rather define one test for one workspace using [langiumGeneratorTest](https://langium-tools.openbrod.org/api/functions/testing.langiumGeneratorTest.html).
|
|
569
592
|
|
|
570
593
|
#### Example Workflow
|
|
571
594
|
|
|
@@ -883,16 +906,16 @@ test("document has no errors, ignoring lexer errors", async () => {
|
|
|
883
906
|
#### API Reference
|
|
884
907
|
|
|
885
908
|
- Matchers:
|
|
886
|
-
- [toHaveNoErrors](https://
|
|
887
|
-
- [toHaveNoIssues](https://
|
|
888
|
-
- [toHaveDocumentIssues](https://
|
|
889
|
-
- [toContainIssue](https://
|
|
909
|
+
- [toHaveNoErrors](https://langium-tools.openbrod.org/api/functions/testing.toHaveNoErrors.html)
|
|
910
|
+
- [toHaveNoIssues](https://langium-tools.openbrod.org/api/functions/testing.toHaveNoIssues.html)
|
|
911
|
+
- [toHaveDocumentIssues](https://langium-tools.openbrod.org/api/functions/testing.toHaveDocumentIssues.html)
|
|
912
|
+
- [toContainIssue](https://langium-tools.openbrod.org/api/functions/testing.toContainIssue.html)
|
|
890
913
|
- Interfaces:
|
|
891
|
-
- [IssueExpectation](https://
|
|
892
|
-
- [IgnoreParameters](https://
|
|
914
|
+
- [IssueExpectation](https://langium-tools.openbrod.org/api/interfaces/testing.IssueExpectation.html)
|
|
915
|
+
- [IgnoreParameters](https://langium-tools.openbrod.org/api/interfaces/testing.IgnoreParameters.html)
|
|
893
916
|
- Enums:
|
|
894
|
-
- [DocumentIssueSeverity](https://
|
|
895
|
-
- [DocumentIssueSource](https://
|
|
917
|
+
- [DocumentIssueSeverity](https://langium-tools.openbrod.org/api/types/base.DocumentIssue.html)
|
|
918
|
+
- [DocumentIssueSource](https://langium-tools.openbrod.org/api/enums/base.DocumentIssueSource.html)
|
|
896
919
|
|
|
897
920
|
#### Notes
|
|
898
921
|
|
|
@@ -903,7 +926,7 @@ test("document has no errors, ignoring lexer errors", async () => {
|
|
|
903
926
|
|
|
904
927
|
In the examples, you might notice the use of a <a href="#adjusted">`adjusted`</a> function to format template strings. This is a utility to clean up multiline strings in tests.
|
|
905
928
|
|
|
906
|
-
See [`adjusted` typedoc](https://
|
|
929
|
+
See [`adjusted` typedoc](https://langium-tools.openbrod.org/api/functions/base.adjusted.html) for more information.
|
|
907
930
|
|
|
908
931
|
#### Conclusion
|
|
909
932
|
|
|
@@ -919,7 +942,7 @@ For more detailed information, refer to the API documentation.
|
|
|
919
942
|
|
|
920
943
|
The adjusted function is a utility for formatting multi-line template strings, ensuring that indentation and line breaks are preserved correctly. This is particularly useful for generating code snippets or structured text where indentation is essential.
|
|
921
944
|
|
|
922
|
-
See [`adjusted` typedoc](https://
|
|
945
|
+
See [`adjusted` typedoc](https://langium-tools.openbrod.org/api/functions/base.adjusted.html) for more information.
|
|
923
946
|
|
|
924
947
|
##### Features
|
|
925
948
|
|
|
@@ -968,7 +991,7 @@ The function returns a formatted string with the following properties:
|
|
|
968
991
|
|
|
969
992
|
##### API Reference
|
|
970
993
|
|
|
971
|
-
- [adjusted](https://
|
|
994
|
+
- [adjusted](https://langium-tools.openbrod.org/api/functions/base.adjusted.html)
|
|
972
995
|
|
|
973
996
|
<p align="right">(<a href="#readme-top">back to top</a>)</p>
|
|
974
997
|
|
|
@@ -1129,7 +1152,7 @@ Don't forget to give the project a star! Thanks again!
|
|
|
1129
1152
|
|
|
1130
1153
|
4. Open a Pull Request
|
|
1131
1154
|
|
|
1132
|
-
5. Release with release
|
|
1155
|
+
5. Release with `npm run release`
|
|
1133
1156
|
|
|
1134
1157
|
### Top contributors:
|
|
1135
1158
|
|
|
@@ -37,6 +37,13 @@ export var DocumentIssueSeverity;
|
|
|
37
37
|
* @param severity - The `DiagnosticSeverity` to map.
|
|
38
38
|
* @returns The corresponding `DocumentIssueSeverity`.
|
|
39
39
|
*/
|
|
40
|
+
/**
|
|
41
|
+
* LSP 3.18 allows a diagnostic message to be {@link MarkupContent}.
|
|
42
|
+
* Langium still emits plain strings. Keep the public issue message a string.
|
|
43
|
+
*/
|
|
44
|
+
function diagnosticMessage(message) {
|
|
45
|
+
return typeof message === 'string' ? message : message.value;
|
|
46
|
+
}
|
|
40
47
|
function mapSeverity(severity) {
|
|
41
48
|
if (!severity) {
|
|
42
49
|
return DocumentIssueSeverity.UNKNOWN;
|
|
@@ -84,7 +91,8 @@ export function getDocumentIssues(document, params) {
|
|
|
84
91
|
}
|
|
85
92
|
if (!skipParserErrors) {
|
|
86
93
|
document.parseResult.parserErrors.forEach((parserError) => {
|
|
87
|
-
//
|
|
94
|
+
// Chevrotain exposes the token before the failure. Inserted EOF tokens use -1
|
|
95
|
+
// for offset, line, and column, which is not a real position.
|
|
88
96
|
const previousToken = parserError === null || parserError === void 0 ? void 0 : parserError.previousToken;
|
|
89
97
|
issues.push({
|
|
90
98
|
source: DocumentIssueSource.PARSER,
|
|
@@ -106,7 +114,7 @@ export function getDocumentIssues(document, params) {
|
|
|
106
114
|
issues.push({
|
|
107
115
|
source: DocumentIssueSource.VALIDATION,
|
|
108
116
|
severity: mapSeverity(diagnostic.severity),
|
|
109
|
-
message: diagnostic.message,
|
|
117
|
+
message: diagnosticMessage(diagnostic.message),
|
|
110
118
|
startLine: diagnostic.range.start.line,
|
|
111
119
|
startColumn: diagnostic.range.start.character,
|
|
112
120
|
endLine: diagnostic.range.end.line,
|
|
@@ -280,17 +288,19 @@ function at(issue) {
|
|
|
280
288
|
return result;
|
|
281
289
|
}
|
|
282
290
|
/**
|
|
283
|
-
* Returns the first
|
|
291
|
+
* Returns the first usable source position from two options.
|
|
292
|
+
* `null`, `undefined`, `NaN`, and negative values are skipped. Chevrotain uses `-1`
|
|
293
|
+
* on inserted tokens (typically EOF) when it has no position.
|
|
284
294
|
*
|
|
285
295
|
* @param a - The first number to consider.
|
|
286
296
|
* @param b - The second number to consider.
|
|
287
|
-
* @returns The first
|
|
297
|
+
* @returns The first usable number, or `undefined` if both are unusable.
|
|
288
298
|
*/
|
|
289
299
|
function firstDefinedNumber(a, b) {
|
|
290
|
-
if (a != null && !Number.isNaN(a)) {
|
|
300
|
+
if (a != null && !Number.isNaN(a) && a >= 0) {
|
|
291
301
|
return a;
|
|
292
302
|
}
|
|
293
|
-
if (b != null && !Number.isNaN(b)) {
|
|
303
|
+
if (b != null && !Number.isNaN(b) && b >= 0) {
|
|
294
304
|
return b;
|
|
295
305
|
}
|
|
296
306
|
return undefined;
|
package/package.json
CHANGED
|
@@ -1,11 +1,12 @@
|
|
|
1
1
|
{
|
|
2
2
|
"name": "langium-tools",
|
|
3
|
-
"version": "0.
|
|
3
|
+
"version": "0.4.0",
|
|
4
4
|
"description": "A set of tools for Langium",
|
|
5
5
|
"type": "module",
|
|
6
6
|
"main": "dist/index.js",
|
|
7
7
|
"types": "dist/index.d.ts",
|
|
8
8
|
"scripts": {
|
|
9
|
+
"release": "npm run test && npm run build && npx release-it",
|
|
9
10
|
"build": "tsc",
|
|
10
11
|
"lint": "eslint 'src/**/*.{ts,tsx}'",
|
|
11
12
|
"lint:fix": "eslint 'src/**/*.{ts,tsx}' --fix",
|
|
@@ -27,11 +28,15 @@
|
|
|
27
28
|
],
|
|
28
29
|
"author": "Boris Brodski",
|
|
29
30
|
"license": "MIT",
|
|
31
|
+
"engines": {
|
|
32
|
+
"node": "^22.22.2 || ^24.15.0 || >=26",
|
|
33
|
+
"npm": ">=10.2.3"
|
|
34
|
+
},
|
|
30
35
|
"dependencies": {
|
|
31
36
|
"@types/node": "^22.0.0",
|
|
32
37
|
"change-case": "^5.4.4",
|
|
33
38
|
"escape-string-regexp": "^5.0.0",
|
|
34
|
-
"langium": "^
|
|
39
|
+
"langium": "^4.4.2"
|
|
35
40
|
},
|
|
36
41
|
"peerDependencies": {
|
|
37
42
|
"vitest": "~2.0.5"
|
|
@@ -42,7 +47,7 @@
|
|
|
42
47
|
"typedoc": "^0.27.0",
|
|
43
48
|
"typedoc-plugin-extras": "^4.0.0",
|
|
44
49
|
"typedoc-plugin-markdown": "^3.0.0",
|
|
45
|
-
"typescript": "
|
|
50
|
+
"typescript": "~5.8.3",
|
|
46
51
|
"typescript-eslint": "^7.18.0",
|
|
47
52
|
"vitest": "~2.0.5"
|
|
48
53
|
},
|
|
@@ -55,5 +60,8 @@
|
|
|
55
60
|
"*.config.*",
|
|
56
61
|
"package-lock.json",
|
|
57
62
|
"*.json"
|
|
58
|
-
]
|
|
63
|
+
],
|
|
64
|
+
"volta": {
|
|
65
|
+
"node": "22.22.2"
|
|
66
|
+
}
|
|
59
67
|
}
|
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
import "./string.js";
|
|
2
2
|
import { LangiumDocument } from "langium";
|
|
3
|
-
import { DiagnosticSeverity } from "vscode-languageserver";
|
|
3
|
+
import { DiagnosticSeverity, MarkupContent } from "vscode-languageserver";
|
|
4
4
|
import type { IToken } from 'chevrotain';
|
|
5
5
|
import { EOL } from "node:os";
|
|
6
6
|
|
|
@@ -131,6 +131,14 @@ export interface GetDocumentIssuesParams {
|
|
|
131
131
|
* @param severity - The `DiagnosticSeverity` to map.
|
|
132
132
|
* @returns The corresponding `DocumentIssueSeverity`.
|
|
133
133
|
*/
|
|
134
|
+
/**
|
|
135
|
+
* LSP 3.18 allows a diagnostic message to be {@link MarkupContent}.
|
|
136
|
+
* Langium still emits plain strings. Keep the public issue message a string.
|
|
137
|
+
*/
|
|
138
|
+
function diagnosticMessage(message: string | MarkupContent): string {
|
|
139
|
+
return typeof message === 'string' ? message : message.value;
|
|
140
|
+
}
|
|
141
|
+
|
|
134
142
|
function mapSeverity(severity?: DiagnosticSeverity): DocumentIssueSeverity {
|
|
135
143
|
if (!severity) {
|
|
136
144
|
return DocumentIssueSeverity.UNKNOWN;
|
|
@@ -188,7 +196,8 @@ export function getDocumentIssues(
|
|
|
188
196
|
|
|
189
197
|
if (!skipParserErrors) {
|
|
190
198
|
document.parseResult.parserErrors.forEach((parserError) => {
|
|
191
|
-
//
|
|
199
|
+
// Chevrotain exposes the token before the failure. Inserted EOF tokens use -1
|
|
200
|
+
// for offset, line, and column, which is not a real position.
|
|
192
201
|
const previousToken = (parserError as unknown as { previousToken: IToken })?.previousToken as IToken;
|
|
193
202
|
issues.push({
|
|
194
203
|
source: DocumentIssueSource.PARSER,
|
|
@@ -212,7 +221,7 @@ export function getDocumentIssues(
|
|
|
212
221
|
issues.push({
|
|
213
222
|
source: DocumentIssueSource.VALIDATION,
|
|
214
223
|
severity: mapSeverity(diagnostic.severity),
|
|
215
|
-
message: diagnostic.message,
|
|
224
|
+
message: diagnosticMessage(diagnostic.message),
|
|
216
225
|
startLine: diagnostic.range.start.line,
|
|
217
226
|
startColumn: diagnostic.range.start.character,
|
|
218
227
|
endLine: diagnostic.range.end.line,
|
|
@@ -403,17 +412,19 @@ function at(issue: DocumentIssue): string {
|
|
|
403
412
|
}
|
|
404
413
|
|
|
405
414
|
/**
|
|
406
|
-
* Returns the first
|
|
415
|
+
* Returns the first usable source position from two options.
|
|
416
|
+
* `null`, `undefined`, `NaN`, and negative values are skipped. Chevrotain uses `-1`
|
|
417
|
+
* on inserted tokens (typically EOF) when it has no position.
|
|
407
418
|
*
|
|
408
419
|
* @param a - The first number to consider.
|
|
409
420
|
* @param b - The second number to consider.
|
|
410
|
-
* @returns The first
|
|
421
|
+
* @returns The first usable number, or `undefined` if both are unusable.
|
|
411
422
|
*/
|
|
412
423
|
function firstDefinedNumber(a?: number, b?: number): number | undefined {
|
|
413
|
-
if (a != null && !Number.isNaN(a)) {
|
|
424
|
+
if (a != null && !Number.isNaN(a) && a >= 0) {
|
|
414
425
|
return a;
|
|
415
426
|
}
|
|
416
|
-
if (b != null && !Number.isNaN(b)) {
|
|
427
|
+
if (b != null && !Number.isNaN(b) && b >= 0) {
|
|
417
428
|
return b;
|
|
418
429
|
}
|
|
419
430
|
return undefined;
|